ATTiny167 Dev Board (Tindie)

[RalphBacon]

Hello Spence

I've ordered one of your ATTiny167 Dev Boards (Tindie Order #348935) because I've reached the end of the road regarding trying to program my blank ATTiny167 chips using ICSP.

I'm using the Arduino as ISP and have never had any issues with that before, programming both bare ATMega328 and ATTiny85 chips just fine.

But this time round all I get is "invalid signature" and either FFFFFFF or 000000 as the signature indicating a wiring error or something. Before you ask, I have all the requisite decoupling caps (which I also use for those other chips) and it would normally be a 30 second exercise to do this! I've tried both the Arduino IDE and AVRDUDE command line with exactly the same result (not unsurprisingly).

As this isn't my first ICSP rodeo I'm not going to spend more time trying to resolve this so I though your dev board would help me by:

a) proving the included Tiny167 can be programmed using Optiboot and ICSP and running the sketch using your excellent TinyCore
b) I can replace the Tiny167 with an IC socket and try and program my chips using ICSP but on your board

If a) works but b) fails then I know it's not my wiring that was wrong but the chips themselves (from AliExpress, so possibly fakes?) Yes, it's happened before with other chips so I won't be surprised.

If you have any further ideas do let me know otherwise we'll see what happens when the board arrives. I'm assuming that for $16 shipping I get it hand delivered on a private jet 😲 in 12 hours? What? What? Oh.

PS How do I pronounce your name? Is it SPENS or SPENSA? and COND or CONDA? If I ever get this chip programmed I'll be showing a project to my viewers and I don't want to get your name wrong! In the UK, your name is not common (typical British understatement).

Best Regards
Ralph Bacon
www.youtube.com/ralphbacon

PPS Tried to send this from the Tindie order page but got an HTTP 405 error. 405? Method not allowed.

[SpenceKonde]

Argh, I hadn't seen this until now because I have been so busy working on the other cores (sometimes I get so many emails overnight that there are new emails pushed off the first page of email,, and in that event I'm generally so overwhelmed that things easily can fall through the cracks)

Dude, $16 shipping is the cheapest international shipping I can get now. The cost to ship out of the USA has gone up more than five-fold. Most of that was before the pandemic too. It started in like 2017. First they required electronic customs clearance, and that increased cost by a factor of more than 2, because now I had to ship to third party in the US who would handle the customs clearance and then apply the proper labels to it.
Then the pandemic hit, and international shipping ground to a halt, and as it came back, the price started rising - we've had several rate increases per year. I am more likely to lose money on shipping than make a tiny profit on the shipping. It's just nuts. And I've tried to get set up with other carriers to compare prices, but UPS rejects all addresses as invalid when I go through stamps.com (which I have to in order to get the commercial rates - otherwise I have to pay the sucker shipping fee, which is about 2.5x the commercial rate, and I havn't managed to get fedex's website to let me sign up, and stamps,com support has never responded to me except an automated message apologizing their slow response - and I'm pretty sure that even commercial rates, fedex and UPS are more expensive than USPS. So - the shipping rates are out of my hands.

The same postal service also loses about 2-3% of international outgoing packages without a trace, and attempting to get information on a package that vanished is no more effective than shouting into a postal dropoff bin. Sometimes packages I ship spend a whole weekend and the monday workday in the back of the mail truck before they start getting processed. Sometimes packages shipped to me show as delivered when they have not been, and they are subsequently delivered several weeks later. Oh, and the offer of shipping insurance is a lie - the procedures to make a claim are hopelessly convoluted. So I just have to expect to reship a couple percent of my outgoing international orders and eat the cost.
And people wonder why US small business has trouble competing on the international market in any sort of manufactured goods; it certainly doesn't help that we're paying $16+ to ship outgoing intl. mail with 2 week+ delivery times. Vendors in other countries however, can offer free shipping on orders worth just a few dollars when shipping to America, and as I understand most other countries. Gee, I can't imagine why we'd be having trouble competing.

First name is pronounced like spencer without the r. Last name of my maternal grandfather, who was from scotland.
And the last name is pronounced condee. Eastern european, and possibly shortened to americanize it because that was very common back in the day, and we find no references to that last name anywhere.

As for the problem at hand - There are a lot of ways to make screw up ISP wiring.

1. the dupont cables could be making poor contact. Dupont line is a consumable - the terminals on the cables are not the spectacular and durable ones made by DuPont, but a chinese clone of Harwin's knockoff of the dupont connector, the M20 - they found a way to use the same header pins, but make the terminal from a single piece of stamped metal. The M20 had most of the problems that the clones have, because the fact is that they're just asking far too much of a bent piece of brass - it has to apply force to keep contact with the pin, and it has to keep the connector in place. The reliable dupont version had a leaf spring in the terminal, which didn't deform, whereas stamped brass does.
2. Is your understanding of which pins to use correct
   MISO is PA2, MOSI PA4, SCK is PA5, and RST is PB7. These are physical pins (starting from 1 and going counterclockwise as is the tradition, on a SOIC or TSSOP package) 3, 7, 8 and 11 respectively. And you need to power both AVcc and Vcc, and both gnd pins need to be grounded, and both of those need a 0.1uF cap between them.

When you get the board, make sure to use the right ISP socket - there's one for if an x7 is installed, and a different one for if an x61 is installed. Seeing as I had those made eons ago and they're only now running out, I guess sharing the board between two parts wasn't such a bad decision. I realized after the fact that the tiny861 doesn't actually qualify to have a board at all - by my internal criteria for making a breakout board, boards which are available in a DIP package do not get a breakout board, unless the DIP version has fewer pins (ex: ATtiny88), but that's beside the point.

Because this was old Atmel, and they didn't belive in standardized anything. There is no evidence in their pinouts that they didn't determine the pinout of each new part with a game of darts, with a diagram of the part taped to the dart board, and darts labeled for each alternate function divided among the design team, except in a rare few cases (the tiny 88 is one - because it was meant to very closely follow the ATmega x8-series. Another is the 841, which was meant to be a replacement for the 84. Except they just couldn't resist moving the int0 pin to a different pin on PORTB.